A Novel Condition-Monitoring Method for Axle-Box Bearings of High-Speed Trains Using Temperature Sensor Signals

Abstract

Axle-box bearing condition monitoring is important for the prognostics and health management (PHM) of high-speed trains. Temperature is an important indicator of the health of rotating machinery. It increases significantly when the machinery reaches a certain stage of failure. However, technologies based on temperature sensor signals for monitoring the status of high-speed train’s axle-box bearings have rarely been reported to date. This paper proposes a novel condition-monitoring method called the “abnormality index model,” which is based on the local outlier factor (LOF) algorithm, for detecting the potential failures in the axle-box bearings of high-speed trains using temperature sensor signals obtained from a wireless transmission device system (WTDS). The proposed method is capable of quantifying the bearing condition as well as reducing any random noise from the external environment that may interfere with the measurement of the temperature signals of the train. WTDS temperature sensor signals for different failure modes were utilized to evaluate the effectiveness of the LOF-based condition-monitoring approach. The results demonstrate that the proposed method can effectively detect temperature signs associated with the health state of the axle-box bearings and uncover the potential failures, even beforehand.